SHADOW BANKING AND MACROECONOMIC EFFICIENCY

(1)

Copenhagen Business School Cand.merc. Finance and Strategic Management

Master Thesis Date of submission: May 15th, 2019

SHADOW BANKING AND MACROECONOMIC EFFICIENCY

The effect of Shadow Banking on key macroeconomic factors

Vanessa Komorowski - 116151 Demetria Parisi - 115485 Supervisor: Søren Ulrik Plesner

Number of pages and characters: 101, 228.805

(2)

Abstract

This thesis investigates the effects that the presence of Shadow Banking activities, in terms of the level of assets belonging to Other Financial Intermediaries (OFIs), can have on one country’s macroeconomic indicators, utilizing a panel data of 26 countries.

In particular, this thesis proposes a two-way fixed effect model, supported by Vector Autoregressive models and co-integration tests to study the influence that OFIs’ assets have on economic efficiency, measured by economic growth (GDP), inflation (CPI), liquidity (broad money), unemployment and the level of real interest rate. Shadow Banking activities do have a significant impact on economic activity, and this relationship results to be overall positive, but turns negative during the post-crisis period, mainly due to tightening of regulatory mechanisms. Moreover, OFIs show a positive effect on monetary indicators by providing significant credit expansion together with commercial banks and this effect is likely to be larger in the pre-crisis period and in developing countries. When turning to the case of inflation, Shadow Banking intermediation has no direct effect, unless, due to weak financial regulations, financial intermediaries have an impact on credit creation, as in the case of developing countries. The causal effect between OFIs and unemployment is weakly significant for most of the countries considered, as well as for the relationship with real interest rate. However, for the latter, in case of a significant impact, the direction of the causality is negative.

(3)

List of Tables

Table 1. Descriptive Statistics – Whole Sample... 36

Table 2. Correlation between selected variables ... 39

Table 3. Regression analysis – Whole sample ... 59

Table 4. Regression analysis – Developed countries ... 60

Table 5. Regression analysis – Developing countries ... 60

Table 6. Regression analysis – Pre-crisis ... 61

Table 7. Regression analysis – Post-crisis ... 62

Table 8. Summary Statistics – selected countries ... 64

Table 9. ADF tests – selected countries ... 65

Table 10. Co-integration tests – selected countries ... 66

Table 11. VAR results – OFIs coefficients ... 67

Table 12. Granger causality test ... 70

Table 13. Robustness check on whole sample - CRVE ... 73 Appendix

(6)

List of Graphs

Graph 1. Interconnectedness between banks and OFIs ... 3

Graph 2. Narrowing down aggregates ... 30

Graph 3. Composition of financial system for sample countries ... 38

Graph 4. Trend in FSB’ variables ... 40

Graph 5. Trend in the mean of OFIs ... 40

Graph 6. Trend in the mean of OFIs for developed vs. developing subsample ... 41

Graph 7. Trend in the mean of OFIs for United States... 42

Graph 8. Trend in the mean of OFIs for China ... 42

Graph 9. Trend in the mean of OFIs for Japan ... 43

Graph 10. Trend in the mean of OFIs for Brazil ... 43

Graph 11. Impulse Response functions ... 69

Graph 12. OFIs and GDP trends – Japan... 79

Graph 13. OFIs and Broad Money trends – US ... 84

Graph 14. OFIs, GDP and Unemployment trends – Japan ... 91

Graph 15. OFIs and Real Interest Rate trends – Brazil ... 94 Appendix

(7)

1

1. Introduction

1.1. Background

More than 10 years after the global financial crisis, which undermined the reputation of international financial markets, the debate on whether the banking sector is enough regulated is still ongoing. What is certain is that conventional banks are more regulated, with more capital and less leverage compared to the years of the crisis. On the other hand, financial regulators are more able to monitor banking activities and to have a clearer overview of their operations.

If this is true for the traditional banking sector, it is not entirely the same for the so-called

“Shadow Banking” activities. They are performed by institutions like, among others, hedge funds, money market funds and real estate funds which all belong to the business of financial intermediation by transferring money from one party to another. While this may sound like the regular routine of a traditional bank, many of these activities within the system of shadow banks are not actively regulated and not subject to the same scrutiny of those of traditional banks.

Moreover, shadow banks are used to engage in activities like maturity and credit transformation which have resulted in issues to be dealt with during the financial crisis, but also in more recent times like during the UK “Brexit” process. On one hand, highly leveraged liquidity and credit transformation were some of the main drivers identified for the financial crisis; on the other, during Brexit, due to the high market uncertainty and fear of asset prices drop, open-ended intermediaries like real estate funds were forced to temporary restrict repayments.

Drawing attention over Shadow Banking activities is important mainly because of their size. In the United States, the Shadow Banking system is almost the same size as the traditional banking sector, while in the European Union it accounts for almost 35% of the size of financial intermediation. This number has been estimated to have been expanded by around 40% in the Eurozone in the years between 2012 and 2016 (Portes, 2019).

Moreover, according to a report from the Financial Stability Board in November 2014, the volume of Shadow Banking activities in the United States has reached 75tn USD.

(8)

2 Since the disruptive effects of Shadow Banking have been revealed in 2007, the role of increased regulation has been crucial in preserving that of Shadow Banking activities.

The Dodd-Frank Wall Street Reform and Consumer Protection Act was signed in July 2010 by President Barack Obama with the aim of regulating Shadow Banking and has been defined as one of the strictest financial reforming acts. Before the approval of the latter, in fact, other financial reforms such as the Financial Modernization Act in 1990 and Commodity Futures Modernization Act in 2000, only led to the further development of financial institutions outside the regulated system (Tang & Wang, 2016).

In an increasingly globalized world, there is the need to believe that the financial instability of one nation might easily affect the one of another and, in turn, trigger a series of responses that will lead to a downward spiral affecting the whole global economy. This is one of the reasons why the need to shed light on Shadow Banking and to regulate it, in order to benefit from its potential, is of extreme importance in modern economics.

This is why the G20 Seoul Summit in November 2010 had, as one of its focus points on the agenda, “strengthening regulation and supervision of Shadow Banking”, asking the Financial Stability Board to intervene on the matter by working together with other international financial and regulatory agencies to reinforce the oversight of the Shadow Banking system (The Seoul Summit Document, 2010).

Moreover, the Basel III regulation, to be implemented in January 2022¹, makes sure that the exposure of banks to Shadow Banking activities is captured. Higher risk weights for unregulated financial intermediaries, risk-sensitive capital requirements for banks investing in equity funds and standards for controlling large capital exposures are all features that raise capital requirements for banks whose activities are linked to Shadow Banking intermediaries.

As depicted by the Financial Stability Board in its 2017 report, and as shown in Graph 1, the interconnectedness between traditional banks Other Financial Intermediaries (OFIs), which include all financial institutions not classified as banks, while it has been declining since the financial crisis, it is still at higher levels than the pre-crisis period, partially

1 Implementation deadline for the disclosure requirements completing Basel III has been postponed to 1 January 2022 (Pillar 3 disclosure requirements - updated framework , 2018)

(9)

3 because of the reliance of banks on investment funds, money-market funds (MMFs) and broker-dealers in funding their activities (2017).

Graph 1. Interconnectedness between banks and OFIs

The still-present relevance of the Shadow Banking system in modern literature and in contemporary economy and financial markets has inspired the following research question.

1.2. Research question

This Master thesis will integrate a broad panel dataset, covering over ten years and twenty countries, to estimate the effects of Shadow Banking on some economic efficiency indicators, alongside country-specific factors. The empirical analysis utilizes data of Shadow Banking levels from the Global Shadow Banking Monitoring Report 2014 issued by the FSB. For macroeconomic variables, we will use mainly the database from the International Monetary Fund, World Bank Database and Federal Reserve.

Our paper focuses on the Shadow Banking sector at the global level to add to the growing research on the topic in recent years. However, the analysis will also attempt to shed some light on the positive effects of the development of Shadow Banking activities on real economy, thus expanding the existent literature, which recently tended to focus on the risks associated with non-financial banks. The analysis further contributes to expand the existing knowledge by utilizing a larger sample of data than previous empirical studies,

(10)

4 while covering a high number of countries with different economies and financial system development levels. Moreover, the analysis will include both panel data models and time series models, in support to the former, to improve the understanding of Shadow Banking development in key economies.

Specifically, the thesis will answer to the following research question:

• What is the effect of Shadow Banking, proxied by Other Financial Institutions (OFIs) assets, on economic efficiency measured by key macroeconomic factors?

In order to cover the main research question, the analysis will be structured around the following sub-questions:

a. What is the effect of Shadow Banking on economic growth, measured by GDP?

b. What is the effect of Shadow Banking on price levels, measured by inflation?

c. What is the effect of Shadow Banking on liquidity, measured by broad money?

d. What is the effect of Shadow Banking on unemployment?

e. What is the effect of Shadow Banking on real Interest Rate?

These questions will be answered by first looking at the whole sample, and by subsequently comparing the effects prior and after the financial crisis (2007-2008), and between developing and developed countries.

By looking at these additional perspectives on the development of Shadow Banking, we aim answering to these following sub-questions:

i. How have the effects of Shadow Banking on economic efficiency changed from before- to after- the crisis?

ii. Are the effects of Shadow Banking development on economic efficiency different when considering developing countries versus developed countries?

This approach will contribute to shed some light on the characteristics of the development of Shadow Banking. By splitting the time frame into two sub-samples based on the financial crisis, we aim at understanding the effect of the policies that took place in

(11)

5 response to the financial crisis. At the same time, we consider the differences in financial systems and economies development levels across countries.

The results of the analysis will therefore also help future research build a discussion around the role of regulations in controlling the development of non-bank financial institutions. For instance, in the case of the activities that are specific to the Shadow Banking system, the objective of the regulation should be to try to prevent systemic crisis and procyclicality of the financial crisis without increasing costs in normal time.

Therefore, it is of great importance to understand how and if to change/improve regulations (Adrian & Ashcraft, 2012).

1.3. Topic delimitation

The literature and economic research on Shadow Banking are extensive and considering how the interpretation of the term “Shadow Banking” has not been clearly defined, there exists many different angles and perspectives that can be chosen for a thorough study on this topic. While countries like China and the United States have been analyzed by many researchers due to the strong presence of the Shadow Banking system in their economies, others have preferred focusing on the Eurozone to estimate the effects of this phenomenon on overall economy.

In this study, in order to have a general perspective on the overall effects of Shadow Banking activities on different economies, we have decided to select a quite large set of countries (around 25), including the United States, China and the Eurozone and to include the largest set of year variables that we could have available. In this regard, we have decided to utilize a dataset provided by the Financial Stability Board which includes 15 years of data, from 2002 to 2016, for 29 world countries.

The Financial Stability Board has a broad definition of Shadow Banking which refers to

“Other Financial Intermediaries”, including all those intermediaries that are considered as non-bank related. We have decided to use OFIs as our main independent variable and to select the definition given by the FSB to delimitate our Shadow Banking characterization.

(12)

6 Furthermore, vast is the research on how the Shadow Banking system has affected the outcomes of the financial crisis (Pozsar, Adrian, Ashcraft, & Boesky, 2013), (Nersisyan

& Wray, 2010) (Bengtsson, 2013) and also broad is the set of studies relating to the need for regulating this phenomenon (Plantin, 2015), (Santos, 2001), (Gorton, Metrick, Shleifer, & Tarullo, 2010).

However, narrow is the set of literature referring to the effects of the Shadow Banking system on real economy indicators, in particular GDP, inflation and liquidity, among others. Hence, we have decided to fill this gap by analyzing the effects of non-bank intermediaries, following the definition of the FSB, on macroeconomic variables, including those mentioned above, for a broad set of countries and for a relatively large time frame.

First, we have decided to narrow our research on specific sub-groups defined both at time and country level. Thus, we have decided to look at the same effect as for the whole sample but for a time specific subgroup, pre- and post- crisis, and for a country subgroup, developed versus developing countries.

Second, our study further aims at narrowing the research gap to specific countries, such as China, the United States, Brazil and Japan. In this regard, with a time series analysis, we look at differences and similarities in the effects of these countries both on a short- and long-term level.

1.4. Thesis structure

This study on the effects of Shadow Banking on macroeconomic factors will be structured by starting from a thorough review of the existing literature on Shadow Banking, from the attempts to define its boundaries to its developments over time. Then, we will review different studies that tried to understand the effects of this phenomenon on real economy.

After having summarized existing studies, we will proceed with our own analysis. First, we define the methodology that we choose to adopt, including data and sample selection.

Then, we present the results of our econometric analysis, discuss the findings and prove their validity with a robustness test. Finally, we conclude our study by describing limitations and future research that can be done to expand the literature.

(13)

7

2. Literature Review

2.1. Shadow Banking 2.1.1. Definition

The term “Shadow Banking” was first used in 2007 by the economist and executive director of the Pacific Investment Management Company (PIMCO) Paul Mcculley during the annual symposium of the Kansans City Federal Reserve Bank.

According to Pozsar (2017), member of the Shadow Banking Colloquium of the Institute for New Economic Thinking (INET), the concept of Shadow Banking is not completely understood by the public and it may be interpreted in different ways by different entities.

In particular, while, for banking institutions, shadow banks might be represented by hedge funds, according to the latter, shadow banks might be denoted by money funds. The reason behind this uncertainty in the summary of what Shadow Banking actually means stands from the lack of a clear definition by its author.

In his speech, Mcculley called “Shadow Banking” the

“alphabet soup of levered up non-bank investment conduits, vehicles and structures.

Unlike regulated real banks, who fund themselves with insured deposits, backstopped by access to the Fed’s discount window, unregulated shadow banks fund themselves with uninsured commercial paper, which may or may not be backstopped by liquidity lines

from real banks.” (McCulley, 2007)

These words have to be placed in the context of the antecedents of the 2007-08 global financial crisis originated in the progressive deterioration of the American financial system. Thus, it is important to mention that the Shadow Banking depiction of Mcculley has a pure U.S. focus and is mainly directed towards the practice of banks of using short term deposits to fund the purchase of long-term assets, normally defined as maturity transformation as well as credit and liquidity transformation. The difference between the short term borrowing of traditional versus shadow banks is that, since the latter are not regulated in the conventional way, they are not able to rely to insured depositors or to the emergency assistance of the Federal Reserve (Kodres, 2013).

In one of his iconic papers on Shadow Banking (authored together with, among others, Tobias Adrian, financial counsellor and director of the IMF’s Monetary and Capital

(14)

8 Markets department), Pozsar tries to specify shadow banks as financial intermediaries that undertake maturity, liquidity and credit transformation activities without access to public sources of liquidity (such as the Fed for the U.S. banking system) and without credit barriers. The web of intermediation created by these exercises has taken the name of “Shadow Banking System” (SBS).

The SBS was funded by collateralized assets and liabilities as a result of the process of securitization, defined as the practice of pooling together certain types of assets to be then repackaged in interest-bearing securities (Jobst, 2008) and exchanged in capital markets.

This form of credit intermediation is often enhanced by liquidity or credit guarantees such as put options provided by the private sector (insurance companies or banks) in the form of wraps, guarantees or credit default swaps (CDS). Some of the means of securitization to intermediate credit within the SBS include asset-backed securities (ABS), asset-backed commercial paper (CP), collateralized debt obligations (CDOs) and repurchase agreements (repos) which were then redistributed in the capital market through Special Purpose Vehicles (SPVs).

While it is theoretically recognized that the SBS makes use of private money to fund its activities, it can also receive implicit or indirect credit quality enhancements from the public sector. Pozsar, Adrian, Ashcraft, & Boesky (2013) define shadow credit intermediation as including all implicitly, indirect or unenhanced activities. Examples of indirectly enhanced credit by the public sector include off-balance-sheet liabilities of banks or off-balance-sheet activities of depository institutions while activities with direct and implicit official enhancement may include those of government sponsored enterprises (GSEs). Finally, money market mutual funds (MMMFs) or bank affiliated hedge funds belong to the group of activities with indirect and implicit official enhancement.

2.1.2. Origins and developments

Pozsar (2008) discusses how the SBS changed the nature of traditional banking, which shifted from a “originate-to-hold” model to a “originate-to-distribute” one and how credit intermediation and risk absorption occurs more into capital markets and more away from banks’ balance sheets. This is mainly because banks do not hold the loans they originate

(15)

9 from private customers, but instead they sell it to broker-dealers who then use dedicated securities to distribute them to investors with different risk appetites.

Adrian & Shin (2009) expand on this by stating that the origins of the SBS in the United States are rooted in the amalgamation of capital markets with the banking sector and the role played by the securitization of assets.

Securitization played a big role in the increased importance of capital markets compared to banking intermediaries in the U.S., mainly because of the enlarged availability of credit supply for institutions engaging in this activity. However, while the original purpose of the securitization procedure was to move credit risk towards financial intermediaries that were more able to absorb losses, the same technique rapidly increased the fragility of the financial system by strenghtening the link between capital markets and banks through the purchase of each other’s securities (Adrian & Shin, 2009).

The system of Shadow Banking intermediation was considered to be a safe one prior to the global financial crisis of 2007-08. This trust was rooted in the inner securities that the SBS provided such as contingent credit lines and tail-risk insurances like comprehensive guarantees. While liquidity shields such as deposit insurances and discount windows were features of traditional banks, credit lines and tail-risk insurances were considered a good substitute from shadow banks, shaping a backstop role for the SBS. As a result of these warranties, originating mainly from capital markets and in particular commercial banks and insurance companies, the SBS managed to implement credit, liquidity and maturity transformation through the issuance of short-term liabilities which were both liquid and highly rated (Pozsar, Adrian, Ashcraft, & Boesky, 2013).

The role played by capital markets in the development of the SBS and the relative rapid increase in the demand for short-term private funds, compared to government public funds has been further analysed by Pozsar (2012) in his paper on the macro view of Shadow Banking. He argued that, in the US and in the years preceeding the financial crisis, institutional cash investors, when looking for safe, short-term and liquid instruments, could choose between government guaranteed instruments (from sovereign financial vehicles, like treasury bills and insured deposits), privately and secured guaranteed instruments (like repos agreements or asset-backed commercial papers from the Shadow Banking system) and unsecured, inguaranteed instruments (like uninsured

(16)

10 deposits and commercial paper from traditional banks). Pozsar demonstrates that, in 2007, the Shadow Banking system developed as a result of a gap between government guaranteed instruments and unsecured instruments. In particular, while the supply of the former was inelastic and insufficient compared to the demand for the same, the latter instruments experienced an inelastic and limited demand since institutional cash investors were reluttant in using uninsured instruments. The instruments provided by shadow banks were appealing and rapidly became the better choice, being liquid, short-term and often guaranteed by AAA rated banks (Pozsar Z. , 2012).

Gorton, Lewellen, and Metrick (2012), on the other hand, argue that, while the U.S.

financial system has changed in the last sixty years with an increase in the importance of the role of the SBS, the share of safe assets to GDP has remained constant. While this holds true, the share of safe assets has shrunk for shadow banks, with the producers of safe assets being governments and not commercial banks. Nevertheless, in their perspective, the collaboration of capital markets, governments and banking sector has resulted in a stable safe-assets ratio.

2.1.3. Risks of private money creation

While the share of safe assets over GDP has been stable, many studies have looked into the increased risks behind the enlarged demand of private money, compared to government backed securities and in the role that the SBS has played in the risk profile of these instruments.

Capital markets represent the basis of the U.S. economy and American capital markets symbolize the most liquid in the world (Bentsen, 2018). In such a market-based economy, the choice of means of funding is closely linked to the variations in the leverage of financial institutions. When a particular financial intermediary grows in its balance sheet assets, this usually entails an increase in trust in the availability of its credit, while a decrease in the same usually anticipates the outbreak of a financial crisis (Adrian & Shin, 2009).

In his working paper, Geanakoplos (2010) builds on the same idea and states that, especially in times of economic crisis, leverage is as important as interest rates in influencing economic variables. According to his theory, supply and demand determine

(17)

11 both equilibrium leverage and interest rate and both interest rate as well as the variation of leverage have an important impact on the price of assets and in the determination of economic bubbles. In turn, he argues that an increase in the perception of safety of a particular institution or instrument leads to an increased leverage that, according to his leverage cycle theory, can result to be unsustainable and play a big role in asset pricing and economic bubble bursts.

While excessive financial leverage can lead to adverse outcomes for the economic system, economists such as Gorton & Metrick (2009) and Stein (2010) still distinguish the positive effects of the safe assets features of the Shadow Banking system. In their studies, however, recognizing the role played by socially excessive leverage, they also suggest the need to implement policies that would control leverage while preserving the creation of liquidity for the financial system. In particular, whilst Gorton and Metrick (2009) focus on the role played by “securitized banking” (the combination of securitization and repo finance) in the financial crisis, Stein (2010) discusses financial stability regulation directed towards private money creation. In his view, if unregulated, private money creation can easily lead to excessive short-term debt and increased vulnerability. On the other hand, he also shows how monetary policy tools can be used to regulate private credit creation.

On a similar note, Gennaioli, Shleifer, & Vishny (2010) support the idea of the need of additional capital and liquidity leading to more stable markets. However, though they acknowledge the positive impact of private safe securities in credit and liquidity creation, they also demonstrate that in most cases, these types of securities inherit an amount of risk that is often neglected. As a result of diversification, insurance and tranching, these securities become attractive to investors being considered as valid substitutes to traditional securities. According to the authors however, because of the neglected risk, they soon become “false positive”, leading to financial instability and lower welfare, even without accounting for excessive leverage.

The literature on the role of private money in credit and liquidity creation originating from the Shadow Banking system further expands towards the many market failures inherent to the securitization process. Securitization and the role of the SBS is deemed by many scholars as one of the main forces driving the financial crisis: as stated by Kodres

(18)

12 (2013) Shadow Banking can be considered as the representations of one of the failures of the financial system and one that brought to the global financial crisis.

In his paper on financial stability, Shin (2009) underlines the importance of the role of credit supply in the securitisation process. While he argues that, before the financial crisis, securitisation was regarded as a way to disperd credit risk, and hence enhance financial stability, he also states that the credit supply had an endogenous aspect in this scenario.

In particular, after the crisis, many blamed securitisation for passing on bad loand to investors. However, in his view, these bad lons were not standing in the hands of unsuspecting investors, but rather in the entire financial system under the balance sheets of financial institutions, which accepted these unsecured instruments to enhance credit supply.

Stein (2010) explores financial stability, with a focus on securitization. In particular, after analyzing the role of asset-backed securities in credit creation and how investors such as hedge and pension funds have utilized these securities, he focuses on the risk sharing role of these instruments and how banks engaging in these activities have circumvented regulatory capital requirements.

Claessens, Pozsar, Ratnovski, & Singh (2012), build on a similar concept and add that Shadow Banking can be seen as a form of “regulatory arbitrage”, defined by Nouy (2017) as the attempt of banks of take advantage of regulatory loopholes and differences in countries’ legislations. In their analysis, Acharyab, Schnabl, & Suarez (2011) prove that the regulatory arbitrage of shadow banks can generate a considerable concentration of systemic risk within the financial sector, which, instead of being transferred away from the regulated banking sector, is strongly correlated with the activities of the SBS.

2.1.4. The role of the SBS in the global financial crisis

As discussed by many scholars, private money creation can often lead to adverse outcomes, especially when the confidence in the underlying warranties starts to weaken.

Since the issuance of the first mortgage-backed-securities in the 1970s, the “originate-to- distribute” model has grown towards more complex structures and the securitization process has become riskier and more opaque. As further explained by Pozser (2008), this evolution was rooted in weak monetary policies and low interest rates which created a

(19)

13 surplus of credit for borrowers and a shortage of interest revenues for investors. This phenomenon was quite evident in particular for mortgage-backed securities, whose warranties were linked to the housing market.

Ashcraft & Schuermann (2008) argue that seven informational frictions, or agency- problems, can arise in the subprime mortgage securitization process, ranging from the complexity of the products offered to the lack of due diligence of asset managers.

Whenever these issues are not properly addressed, then excessive lowering of standards and increasingly aggressive complexity of securities can easily lead to a subprime crisis of the financial system.

Pozsar, Adrian, Ashcraft, & Boesky (2013) discuss the downfall of the Shadow Banking system and how it can be attributed to the fact that many of the parties involved, including credit rating agencies, regulators, investors and risk managers, did not fully considered the correlation between the aggregate risk and asset prices for these instruments, which becomes much more strong in extreme environments compared to normal times. Coval, Jurek, & Stafford (2009) further expand on this concept by underlying the misconception about the safety of these assets. In particular, the expansion of the securitization procedure and its extraordinary success was rooted in the conception that the prioritization capital structure used in creating tranches of claims was creating safe assets from risky collateral.

The financial crisis later revealed how this conception was wrong and how these instruments were actually far more risky than anticipated.

The outcome of the overestimation of the value of private credit enhancement was an excess of credit supply and the resulting underpricing of public sector liquidity and credit money increased risk-taking attitudes. Once the risk of these securities started to become umbearable and once the subprime financial crisis bursted, the SBS started to rapidly decline. After the failure of Lehman Brothers in 2008, many of the activities undertaken by shadow banks collapsed and the remaining actions were only saved by the creation of official liquidity guarantees that substituted the private sector, such as emergency liquidity facilities and guarantee schemes by government agencies (Pozsar, Adrian, Ashcraft, & Boesky, 2013).

(20)

14 2.1.5. Modern Shadow Banking

According to the Financial Stability Board (FSB)’ annual monitoring of Shadow Banking activities in 2018, Shadow Banking accounts today for 13% of total global financial assets. After the FSB has been able to include data from China and Luxembourg, the amount of Shadow Banking assets grows up to more than $45 trillion (Binham, 2018).

In 2017, the FSB has claimed to have taimed the “toxic” bits of Shadow Banking that have led to the development of the financial crisis thanks to many of the financial regulations and reforms put in place by G20 countries. According to Mark Carney, chairman of the FSB and of the Bank of England, what is left of Shadow Banking activities is mainly market-based finance. He argues that the regulation implemented has helped both to eliminate the harmful side of Shadow Banking as well as keeping the its valuable parts, making large banks more strong and more compliant to stress tests and capital requirements (Stafford, 2017).

2.2. Shadow Banking and its effects on real economy

Many scholars have focused on defining the boundaries of Shadow Banking while many others have investigated how securitization of assets and the role of capital markets have affected the developments of the global financial crisis. Some others, instead, have looked at how the downfall of the SBS has affected macroeconomic efficiency and at how it has impacted real economy. However, while the majority of the studies focuses on how changes in the dynamics of the financial economy, including the transition towards a

“originate-to-distribute” banking sector, have had an influence on the evolution of the SBS, not many look at the effects that Shadow Banking has had on real economy.

Before the development of what is today called Shadow Banking, some studies already highlighted how bank failures impact overall economy. Particularly, Bernanke (1983), concentrates on the credit related non-monetary aspect of the financial sector and indicates how a financial crisis, in particular one involving bank failures, negatively affects the efficiency of credit allocation, leading to higher costs and lower availability of supply of money, in turn affecting aggregate demand.

(21)

15 Supporting this theory is Ashcraft (2003) which points at how healthy bank failures have significant and permanent effects on real economic activity, which are usually underestimated compared to when traditional banks are in distress.

The reputation of the SBS has been questioned especially after the global financial crisis has partially blamed it for the resulting financial instability and highlighted its potential credit risk and its lack of adequate safety measures compared to the traditional banking system. This has led to the growth of the Shadow Banking literature focusing on its negative aspects and on the negative outcomes that it transfers to overall economy.

However, some economists also look at the bright side of the Shadow Banking system and at how its original purpose of increasing credit intermediation can still bring positive outcomes to the real economy.

Moreira and Savov (2014) discuss this duality in effects by focusing on the tradeoff between economic growth and financial stability. While Shadow Banking boosts economic growth and alleviates the shortage of safe assets by transforming risky and illiquid assets in securities that are traded liquid, it also creates fragility and can aggravate the effects of a crisis when its inherent liquidity vanishes.

Advocates of the latter view define Shadow Banking as a way to evade capital requirements or regulatory arbitrage (Buchak, Matvos, Piskorski, & Seru, 2017) and as a way to hide risky securities under a façade of safe assets to be sold to unsuspecting investors. On the other hand, the safe-asset view, pictures the SBS as a value-adding way to create money-like liquid securities (Gorton, Lewellen, & Metrick, 2012).

In a world where safe and trustable collateral is limited, even for traditional banks utilizing deposits or short-term government obligations, providers of yet-not-fully-stable liquidity instruments like shadow banks can represent the next-best alternative. In their paper, Moreira and Savov (2014), show how the increased level of liquidity provided by shadow banks reduces the cost of capital for firms while increasing investment and economic growth. Furthermore, they show how, in times where the level of credit intermediation through Shadow Banking is quite extensive, Shadow Banking instruments are able to fund riskier but also more productive investments, moving up the risk-return frontier. However, they also point out at the fact that this process creates financial fragility because, when there is a Shadow Banking boom, the first signs of trouble in terms of

(22)

16 liquidity and security guarantees can lead to a cascade of adverse events. When the SBS fails and the means of funding provided by shadow intermediaries are no longer available, liquidity creation requires more collateral, leading to an increase in discount rates and a downgrade in asset prices, followed by higher margins. This cycle causes an intensified collateral shortage and liquidity crunch, which will soon induce a recession. This downward cycle has induced economists to define the Shadow Banking system as a

“negative externality” on the rest of the economy.

On the positive spectrum of the effects of the SBS on real economy, Ruan (2018) studies the repercussions of the recent extraordinary growth of Shadow Banking in China, focusing on the role of non-financial firms in supplying credit to other firms in the form of entrusted loans within the Shadow Banking sector, acting as “surrogate intermediaries”. His results show that entrusted lending reduces the gap in bank loan supply caused by regulation and does not undermine financial stability. This is due to the fact that entrusted lenders tend to rely on existing cash to make loans and not on external finance, reducing the risk of affecting financial stability.

The Shadow Banking system in China was also analyzed by Sun and Jia (2018) who distinguish between banks’ shadow and traditional Shadow Banking, where the former refers to bank’s credit creation through accounting items generating liabilities from assets.

They prove that, in China, banks’ shadow represents the largest proportion within the SBS and it has strong influence on economic growth and social wealth distribution.

Bianco (2015) develops an accounting model that explores the influence of financial markets on real economy and in the activity of banks of doing “finance through money creation”. He demonstrates that securitization, involving Shadow Banking activities, does not have an effect on the financial sustainability of growth per se, but, nevertheless, it has to be regulated to check for predatory lending.

The relationship between regulatory agencies and shadow banks is analyzed by Benink (2016). In particular, he points at the increased risk of encouraging Shadow Banking activities when creating a capital markets union (CMU), under which the financial system is transformed into one where funding is directed to firms and households through securities and non-bank markets. On the other hand, the increased complexity in regulations and the high level of non-performing-loans (NPLs) from the traditional

(23)

17 banking sector has encouraged the shift towards the less regulated and more opaque system of Shadow Banking.

On the other hand, Meeks, Nelson and Alessandri (2013) state that, following a financial shock, regulatory policies that are solely aimed at stabilizing the securitization market are relatively ineffective. In their dynamic macroeconomic model, the traditional commercial banking sector and the banking sector interact through the securitization market.

Furthermore, their findings show that, while securitizing otherwise unfit loans is good for aggregate activity and credit supply, when the SBS becomes high leveraged then the economy becomes too vulnerable to negative externalities such as liquidity shocks.

2.3. Empirical Evidence

The existing literature about the relationship between economic efficiency and the financial system consists of a broad amount of research studies which, while on one hand attempt to identify the macroeconomic factors that influence the development of the financial system, on the other hand attempt to explain the effects of the financial system development over economic growth. Most of these studies are mainly based on the context of financial/banking crises.

Due to the lack of statistics on the Shadow Banking activity and the ambiguity on the definition of Shadow Banking (FSB, Global Financial Stability Report, 2014), there have not been many researches that provide empirical evidence on its impacts at a global level.

Quantitative research applied to Shadow Banking, in particular in the Euro area, is still at an initial stage. This lack of empirical models on the influence exerted by Shadow Banking is also due to the relative short time series of the components of Shadow Banking system, and the heterogeneous nature of entities, activities and instruments that compose Shadow Banking. Most studies address only the conceptual perspective of the various aspects of Shadow Banking activity and its prospects (Barbu, Boitan, & Cioaca, 2016).

2.3.1. The drivers of Shadow Banking

The main existing empirical studies on non-bank finance have focused on the drivers behind the rise of Shadow Banking over the years (Duca, 2016). The International Monetary Fund report on the Global Financial Stability (FSB, Global Financial Stability Report, 2014) tries to overcome the shortcomings of the existing data by using the flow

(24)

18 of funds measure as a proxy for the Shadow Banking system. The econometric analysis identifies different factors that affect the rise of Shadow Banking: bank regulation, liquidity conditions, financial development, GDP growth and the banking sector size.

Regarding the impact of regulation in specific, D. Enste (2010) shows that regulations are the main causes for the size of shadow economy based on an analysis covering 25 OECD countries.

F. Malatesta, S. Masciantonio & A. Zaghini (2016) conduct an analysis of the main determinants of the loans from shadow banks in the Euro Area. By applying the model on both funding sources from SB and from banks, the study concludes that shadow banks have been influenced to a larger extent by demand-side factors as the macroeconomic conditions of the area. Similarly, T. Barbu, I. Boitan and S. Cioaca (2016) find that Shadow Banking is negatively influenced by short term interest rates and by the ratio of investment funds’ assets in GDP, and positively determined by stock index dynamics and long term interest rates.

A debated research question has been whether monetary policy was an important driver of financial intermediaries’ balance sheets dynamics in the years before the financial crisis. Nelson, Pinter & Theodoridis (2015), investigate the common assumption that monetary policy should have been tighter to usefully “get in all the cracks” of the financial sector in a uniform way. Their empirical results using a vector autoregressive model with monetary policy shocks, commercial banks and shadow banks variations on assets growth in the US, shows that the effects are different. In general, less than 10% of the variation in the quarterly asset growth of commercial and shadow banks over the period 1966-2007 was accounted for by monetary shocks. In the period since 2001, unexpectedly loose monetary policy contributed little to the balance sheet expansion of US financial intermediaries.

On the other side, Chen, Ren, & Zha (2018) empirically argued that contractionary monetary policy during 2009-2015 in China caused Shadow Banking products to rise rapidly. Using a dynamic panel vector autoregressive (VAR) model, the analysis provides evidence of the impact of monetary policy, which gives non-state banks a strong incentive to take advantage of the lax regulatory environment by increasing Shadow Banking activities and by bringing Shadow Banking products into a special investment category

(25)

19 on the asset side of their balance sheets. The estimation indicates that in response to a one-standard-deviation fall of exogenous money supply growth, bank loans fall persistently. The estimated dynamics are statistically significant and the estimated response of Shadow Banking activities is positive over time. As a result of the Shadow Banking rise, the effectiveness of monetary policy on the total credit level in the banking system was hampered.

Other studies consider also some non-economic-related factors that drive Shadow Banking. An interesting analysis by Adriana Davidescu (2015) shows the existence of unidirectional causality between employment rate of older workers (proxy for active ageing) and shadow economy. This result suggests that the unofficial sector represents a social buffer for older workers who have lower labor market opportunities. G. Buchak, G. Matvos, T. Piskorski, and A. Seru (2017) examine the role of technology in the conforming mortgage market in the US, in the period between 2007 and 2015. In particular, the analysis recognizes three types of lenders competing for mortgage borrowers: banks, non-fintech shadow banks and fintech shadow banks. The study proposes a calibrated model that decompose the contribution of different factors to the growth of shadow bank. Interestingly, technology alone is responsible for approximately 78% of gains of fintech firms, and 35% of shadow bank growth overall.

2.3.2. The impacts of Shadow Banking

In recent years, a number of empirical studies emerged, exploring the economic impacts of Shadow Banking by focusing on its different components and/or on different countries/regions. The focus on the effects of Shadow Banking on the financial market and contagious problems to the whole system is widespread across the literature (Xuan

& Quoc, 2016).

Pozsar et al (2010); Adrian and Ashcraft (2012) and Adrian, Ashcraft, Cetorelli (2013) review the rapidly growing literature on Shadow Banking and provide a conceptual framework for its effects. They argue that interconnections of shadow banks with other financial institutions create sources of systemic risk for the broader financial system, and that it negatively impacts financial system stability due to its regulatory arbitrage, neglected risks, funding fragilities as well as high leverage and agency problem. Y.

Maeno, K. Nishiguchi, S. Morinaga & H. Matsushima (2014) develop a systemic risk

(26)

20 model to investigate the impact of shadow banks on the severity of financial contagion and concluded that financial contagion from shadow banks causes the bankruptcies of regulated banks when the market share of Shadow Banking is at 10% to 50%.

The current literature presents considerable studies on Shadow Banking economic costs and benefits in China and other Asian countries. The “non-bank financial intermediaries”

(NBFIs) and “Other Financial Institutions” (FSB, Global Financial Stability Report, 2014) in Asia provide alternative sources of financing to the underserved market segments in the economy. According to the Global Shadow Banking Monitoring Report (FSB, March 2018), Chinese shadow Banking, proxied by the assets of the other financial institutions, has reached 9.5 USD trillion, which accounts for 85% of GDP. The products involved are much simpler: for example, the main Shadow Banking products in China are trust products, wealth management products and entrusted loans (Chao, Shan, & Zhang, 2017).

For instance, using China’s data, C-C Chao, M. Shan, and J. Zhang (2017) examine the impacts of Shadow Banking on real output in the long and in the short run. The empirical results indicate that in the long run the expansion of Shadow Banking can result in the widening of the wage gap between skilled and unskilled labor, and thus reduce the real output of the economy, while in the short run the development of the Shadow Banking sector can increase an economy’s social welfare. The empirical results conclude that Shadow Banking in dual economies and developing countries positively impact economic growth and economic output. C-C Chao, M. Shan, and J. Zhang (2017) find that a 1%

increase in Shadow Banking raises economic growth by 0.48% in the short run, using a Fixed Effect regression with Shadow Banking, financial intermediary indicators and other control variables. Similarly, other studies such as Zhang and Peng (2014), Jerome (2015) and Li and Li (2015), empirically show a positive relationship between non-bank financial intermediaries in developing countries and real economy, using panel VAR models and impulse response function analysis on China’s datasets.

Using Johansen test (co-integration) and vector error correction models, R. Rateiwa & M.

Aziakpono (2017) empirically test the existence of a long-run equilibrium relationship between economic growth and the development of non-bank financial institutions, and the causality thereof, in Africa’s three largest economies, namely Egypt, Nigeria and

(27)

21 South Africa, over the period 1971–2013. The results show that the relationship between NBFIs development and economic growth in Egypt is positive and significant, and predominantly bidirectional. In South Africa, the causality is positive and significant and predominantly runs from NBFI development to economic growth, implying a supply- leading phenomenon. In Nigeria, the results are weak and mixed. Overall the authors argue that the empirical evidence suggests a stronger impact of NBFIs development on economic growth in countries with a more developed financial system.

Other two studies address the relationship between Shadow Banking activities and economic growth in Nigeria, Ndugbu et al. (2015); Osuala & Odunze (2014), covering the period 1992-2012 and using an ordinary least squares method and an autoregressive distributive lag (ARDL) model, respectively. Both studies conclude that there exists a positive relationship between assets of insurance companies and economic growth.

On the other hand, the rise of Shadow Banking in China seems to have two-sided impacts on the economy, which includes risk factors that must be considered. Based on vector auto-regression model (sVAR), Zheng & Hui (2014) conduct an empirical analysis on the influence of Shadow Banking on monetary policy by using monthly data between 2003 and 2013. The analysis shows that the development of the Shadow Banking system affects the monetary policy transmission mechanism such as the conduction effect of credit and interest rate, which makes it more difficult for the central bank to make monetary policy than before. In the U.S., José Luiz Rossi Júnior (2013) applies a macroeconomic model both in-sample and out-of-sample, and finds that changes in funding liquidity of U.S.

financial intermediaries impact exchange rates around the globe. In particular, the long- term interest rate and risk-taking indicators have robust in sample and out-of-sample predictive power with respect to exchange rates.

Liang & Reichert (2012) argue that if Non-Bank Financial Institutions are not properly regulated, they allow excessive risk appetite, which may have disastrous consequences for both the financial sector and the real economy. Their empirical analysis finds that NBFIs growth has a statistically significant negative impact on economic growth, when using financial data from the recent financial crisis period on a cross-country dataset for both emerging and advanced countries.

(28)

22 Zou, Pang, & Zhu (2013)’s empirical analysis with co-integration test and error correction model demonstrates that overgrowth of Shadow Banking will definitely increase financial fragility. Shadow Banking plays a positive role on the premise of ideal system conditions including comprehensive laws and regulations such as instructive reform mechanism, industry self-discipline regulation, market supervision mechanism and information revelation system, etc.

Building on the impact of monetary policy on the rise of Shadow Banking in China, Chen, Ren, & Zha (2018) further analyze the consequences of this Shadow Banking activity.

Through a quarterly dynamic panel model with Chinese banks and financial intermediaries’ data, the study identifies that the rise of Shadow Banking products makes monetary policy ineffective on the total credit because the Shadow Banking loans rise offsets the decline of bank loans.

The same reflection is supported by Z. Pozsar (2014), who studies the global funding flows and argues that the monetary aggregates and the financial accounts of the United States do not adequately reflect the institutional realities of the modern financial ecosystem, and should be updated to allow policymakers to better analyze and monitor the Shadow Banking system and its potential contributions to financial instability.

Few studies focus on the effect of Shadow Banking activities on the Shadow Banking institutions itself. Xuan & Quoc (2016) focus on Vietnamese securities firms engaging in Shadow Banking activities. Through a pooled regression and a panel regression, the analysis shows that the size of Shadow Banking activities of the securities firms is negatively related to their overall financial conditions. Credit services have a negative impact on asset quality, capital adequacy and liquidity. The effect of Shadow Banking activities on the profitability of securities companies is positive, but this effect is not significant when other control variables, such as age and number of services, are included.

J. Tang & Y. Wang (2016) study the effect of the share of net fees and commissions income to total revenue on the return and risk-adjusted return of Chinese commercial banks. The empirical analysis considers annual fiscal data from 43 Chinese commercial banks and applies a multivariate regression analysis. The dependent variables chosen are the Return on Assets and the Sharpe Ratio. The results show that Shadow Banking activities in commercial banks are associated with higher returns and risk. The Shadow

(29)

23 Banking independent variable has indeed a positive and significant impact both on ROA and on the Sharpe ratio for the selected sample of commercial banks.

2.4. Theoretical background and hypotheses

Considering the research questions and the scope of this study, we draw our hypotheses from both the macroeconomics theory and the financial literature. There is indeed a general acceptance of the fact that there exists a significant direct relationship between financial markets and macroeconomic variables. Economists are now trying hard to incorporate financial factors into standard macroeconomic models (Borio, 2014).

Based on the findings of previous studies and empirical researches on the impacts of Shadow Banking, we suggest possible outcomes for our empirical analysis, summarized by six hypotheses, as presented below.

2.4.1. Role of financial sector in economic growth

At the beginning of the XX century, Schumpeter argued that the services provided by financial intermediaries, mobilizing savings, evaluating projects, managing risk, monitoring managers, and facilitating financial transactions, are essential for technological innovation and economic development (Schumpeter, 1911).

Later on, M. O. Odedokum (1996) developed a neoclassical growth model based on a framework that recognizes the existence of external effects of financial intermediaries on the real sectors and the effects of the financial sector effects on the efficiency of factors of production employed in these real sectors. The study empirically tests the model on a sample of 71 developing countries and the main results show that financial intermediation promotes economic growth in most of the countries (about 85% of the 71 countries). In practical terms, financial intermediation has an equal impact of export growth and capital formation-GDP ratio and a superior impact to labor force growth, in promoting economic growth. Finally, the growth-promoting effects of financial intermediation are more predominant in the low-income than in the high-income developing countries.

The Neoclassical economic growth theory is based on the assumption of constant returns to scale in production technologies. On the other hand, the endogenous economic growth model allows that technological changes over time have not been equally transmitted in developing countries (Gulzar, 2018). Besides integrating macroeconomic growth models

(30)

24 with technology growth and knowledge factors, the models emphasize the role of the financial sector.

For instance, the basic endogenous growth model, the “AK” model, considers aggregate output as a linear function of aggregate capital stock. This reduced form can be expanded into savings and investment, including the role of financial intermediaries. The model so expressed reveals that financial intermediation can affect economic growth by acting on the saving rate, on the fraction of saving channeled to investment, or on the social marginal productivity of investment (Pagano, 1993).

The relevant literature, the empirical evidence and the main macroeconomic theories, provide the basis for the following sub-hypotheses:

→Hypothesis 1. Shadow Banking, defined by Other Financial Intermediaries (OFIs) has a significant impact on GDP, our proxy for economic growth, and this relationship might be stronger in developing countries;

→Hypothesis 2. The overall impact is positive, but the existence of the financial crisis has a negative impact on the relationship.

According to both the neoclassical economic growth model and the endogenous growth model, the financial sector enters the equation of economic growth.

To narrow down the impact of financial intermediaries, Hypothesis 1 builds on the findings of several studies analyzing the relationship between Shadow Banking and economic growth as outlined in Section 2.2. These studies emphasize the role of Shadow Banking and financial intermediaries in the determination of economic output. Several empirical researches found stronger significant results in developing countries, such as China, Nigeria, South Africa and others (Section 2.3.2).

The sign of the expected impact of Shadow Banking on GDP has been a discussed topic in the main literature. Moreira and Savov (2014) explain the trade-off between economic growth and increase systematic risk as a consequence of the rise of Shadow Banking activities. While many scholars argue that these activities are harmful for the economic output, China and several developing countries represent strong positive examples for financial intermediaries as Shadow Banking has a positive and significant impact of economic growth in these countries.

(31)

25 Thus, our Hypothesis 2 suggests that the positive impact of Shadow Banking on the main developing countries might result in an overall positive sign for the variable Shadow Banking. This hypothesis is conditional to the effect of financial crisis in the sample.

2.4.2. Financial sector, inflation and broad money

According to the most relevant macroeconomic theories on inflation, growth of money supply is the broadly identified cause of inflation in developed countries. In developing countries, in contrast, inflation is not a purely monetary phenomenon. For instance, factors typically related to fiscal imbalances such as higher money growth and exchange rate depreciation arising from a balance of payments crisis drive inflation in developing countries (Totonchi, 2011).

Keynesian theory (1936), from John Maynard Keynes (1883-1946), underlines that the increase in aggregate demand is the source of the so-called demand-pull inflation. The aggregate demand comprises consumption, investment and government expenditure.

According to demand-pull inflation theory of Keynes, a policy that causes a decrease in each component of total demand is effective in reducing the reduction of pressure of demand and inflation. One of the reductions in government expenditure is tax increase and controlling the volume of money, alone or together, can be effective in reducing effective demand and inflation control.

The following are two distinct relationships between the inflation rate and economic growth.

The first is the short-run Phillips curve relationship, named after William Phillips (1958) who argued that there is a stable negative relation between the level of unemployment and the rate of change of wage. The connection with inflation was subsequently introduced by Milton Friedman (1977) and Edmund Phelps (1967). In the short-run, a higher level of inflation is positively related to a higher output growth and the intuitive reason is that in the short-run economic expansions will often create demand pressures that lead to inflation (Totonchi, 2011).

The second relationship suggests that inflation has a negative impact on growth in the long run that is due to high inflation episodes; the threshold for an inflation effect on growth may be as high as 40 percent per year (Rousseau & Wachtel, 2000). The negative

(32)

26 impact of inflation on growth in the long run can be due to direct effects and to indirect effects through the financial sector.

Relevant literature focusing on the relationship between inflation and financial development agrees on the unidirectional negative causality between inflation and both banking sector development and equity market activity (H.Boyd, Levine, & D.Smith, 2001).

But according to Monetarist theory, inflation is essentially a monetary phenomenon in the sense that a continuous rise in the general price level is due to the rate of expansion in money supply far in excess of the money actually demanded by economic units (Obi

& Uzodigwe, 2015). There is a well-established long-run empirical relationship between broad money growth and inflation across a variety of countries and monetary regimes. At the same time, J. Cohen (1954) proves the contribution to money supply by financial intermediaries such as life insurance companies, savings and loan associations, and mutual savings banks. Moreover, S. Berry, R. Harrison, R. Thomas & I. de Weymarn (2007) argue that innovations in the financial sector are another key source of changes in money growth.

→Hypothesis 3. As Shadow Banking, measured by OFIs, increases overall liquidity, it also has a positive effect on broad money.

→Hypothesis 4. Overall, the relationship between inflation and OFIs is unidirectional from the first to the latter, and thus the impact of Shadow Banking on inflation is insignificant. However, assuming that Shadow Banking increases broad money and reduces monetary policy effectiveness, then inflation might be positively affected.

2.4.3. Financial sector and unemployment

Classical labor theory is not well suited to consider unemployment, as according to this framework the amount of labor that workers supply is exactly equal to the amount of labor demanded by firms at the equilibrium wage (Fitzgerald, 1998).

As previously mentioned, the most relevant theory on the relationships of unemployment is represented by the Phillips curve, where the lower the rate of unemployment, the higher the rate of inflation.

SHADOW BANKING AND MACROECONOMIC EFFICIENCY

Copenhagen Business School Cand.merc. Finance and Strategic Management